Embed Size (px)
Citation preview
Coordination of HVDC interconnections
Spyros Chatzivasileiadis
multiDC Projectwww.multi-dc.eu
multiDC2
Innovative Methods for Optimal Operation of Multiple HVDC Connections and Grids
• Innovation Fund Denmark Grand Solutions
• Partners:– Two neighboring TSOs:
Energinet, Svenska kraftnät– Three universities:
DTU, KTH, Univ. of Liege– One major manufacturer: ABB– Advisory Board: RTE, Nordic RSCI
• 4.2 million USD• 4 years; Start May 1, 2017
Three main drivers
• 100% renewables– Varying inertia systems– Uncertainty
• 100% inverter-connected devices– How is stability and operation
affected?– How to model them?
• HVDC Grids
3
Kriegers Flak• Denmark – Germany: AC+HVDC
• First interconnection in the world that integrates off-shore wind farms along its path
• 400 MW Back-to-Back HVDC
• Wind Farm Kriegers Flak (DK) : 600 MW
• Wind Farm Baltic (DE) : 336 MW
• HVDC Master Controller to:– Control voltage– Avoid overloadings– Ensure market outcome by mitigating wind
forecast errors
4
North Sea Wind Power Hub• Construction of island(s) in the middle
of the North Sea
• Integration of up to 150 GW of off-shore wind farms
• HVDC interconnections to Denmark, Germany, the Netherlands, UK, Great Britain, Norway, Belgium
• Coupling the energy markets
• Agreement between Denmark, Germany, and the Netherlands already signed (2017)
5
Grid Connection Options for Offshore Wind
6
AC platform
DC platform
Hub as an island
Close to shore
Far from shoreLarge scale far from shore
> 4 GW wind farms
> 80 km offshore
Information from Peter Larsen and Fitim Kryezi, Energinetand from www.northseawindpowerhub.eu
• Modular approach• Each island: up to 30 GW
Vision: 150 GW in North Sea• First step: 12 GW
7
• Far-shore becomes near-shore• Distribution point for different
countries• 2030 and beyond
Information from Peter Larsen and Fitim Kryezi, Energinetand from www.northseawindpowerhub.eu
How will the wind farms be connected to the island?
How will the island be connected to the mainland?
These are still open questions
8
A connection possibility9
Information from Peter Larsen and Fitim Kryezi, Energinetand from www.northseawindpowerhub.euFigures courtesy of Ørsted A/S, Siemens, and Northseawindpowerhub
AC
Mai
nlan
d
AC: Low-frequency AC
HVDC: Point-to-point HVDC
10
ACAC
UK
Netherlands Germany
Denmark
Norway
AC Ring on the island
Challenges and Opportunities• Zero-inertia AC Ring
– Fast transients
• Coordination of the control of the VSC converters– Grid-forming shared among the
converters?– Dealing with failures (N-1)
• Sharing wind power among the countries– Ownership of wind farms– Do we need to adapt the market
structures?
11
multiDC: Addressing current challenges while preparing
for the North Sea Wind Power Hub
12
The three pillars of multiDC13
Robust frequency control of varying inertia systems
Coordinated control of AC/DC systems
Market integration of meshed HVDC connections
Implementation at PowerlabDK
From Current Challenges to the North Sea Wind Power Hub (NSWPH)
14
Robust frequency control of varying inertia systems
Coordinated control of AC/DC systems
Market integration of meshed HVDC connections
Zero-inertia AC grids
Coordinated control of grid-forming HVDC converters
Integration of the NSWPH to the European Market
Current Work By the end of the project
Robust Frequency Control for Varying Inertia Systems
• Decreasing inertia should improve the damping ratio– di/mi describes how fast a frequency deviation is brought back to
equilibrium
• Decreasing inertia should increase ROCOF– Disturbances are scaled by 1/mi
– With low inertia the rotor speed becomes more vulnerable to shocks
15
𝑚𝑖�̇�𝑖 = −𝑑𝑖𝜔𝑖 + 𝑝𝑚𝑒𝑐ℎ,𝑖 − 𝑝𝑒𝑙,𝑖
• mi and di vary depending on the RES infeed – more RES infeed less conventional generation lower inertia– less RES infeed more conventional generation higher inertia
• H∞ optimal control– Minimizes the maximum singular
value of the closed-loop system
• Robust frequency control– Attenuate the gain at higher
frequencies, resulting to lower ROCOF and lower maximum frequency deviation
– Adds some slight damping to electromechanical oscillations
16
Robust Frequency Control for Varying Inertia Systems
0 2 4 6 8 10Time t (s)
-0.01
0
0.01
0.02
Initial inertia level
Lower inertia level
Robust control
Misyris, Chatzivasileiadis, Weckesser, Robust FrequencyControl for Varying Inertia Power Systems, accepted at ISGTEurope 2018, link to paper
Robust Frequency Control for Varying Inertia Systems – Future Steps
• From low-inertia to zero-inertia- Zero inertia Coupling between active and reactive control in
the absence of a stiff frequency and voltage
17
North Sea Wind Power Hub as a test case
Coordinated control of Multi-Area AC/DC systems
• Focus on Emergency Power Control (EPC) mechanisms and sharing of reserves between asynchronous systems
• Currently, EPC in Nordics works as follows:– If f < threshold then transfer = xx MW
• Goal: move from stepwise-triggers to droop-frequency control
– Transmitted power is continuous and linearly dependent on the frequency deviation
18
19
EPC: Trigger (existing) vs Droop (proposed)
Obradovic, Ghandhari, Eriksson, Assessment and Design ofFrequency Containment Reserves with HVDC interconnections,accepted at NAPS 2018, link to paper
• Trigger: power continues to get transferred even if ROCOF becomes positive– This power does not help reduce the
frequency nadir• Droop: for any inertia level, the required
power is less than in the “trigger” EPC
Droop Gain of HVDC (MW/Hz)
Inertia levels
20
EPC: Trigger (existing) vs Droop (proposed)
Obradovic, Ghandhari, Eriksson, Assessment and Design ofFrequency Containment Reserves with HVDC interconnections,accepted at NAPS 2018, link
• Trigger: power continues to get transferred even if ROCOF becomes positive– This power does not help reduce the
frequency nadir• Droop: for any inertia level, the required
power is less than in the “trigger” EPC
Droop Gain of HVDC (MW/Hz)
Inertia levels
Market Integration of HVDC21
• HVDC interconnectors are usually longer than AC interconnections
• HVDC losses are not negligible
• If price difference between areas is small, TSOs cannot recover the cost of HVDC losses– Cost of the losses higher than potential
revenue
• Introduction of an HVDC loss factor in market clearing*
*Fingrid, Energinet, Statnett, Svenska kraftnät, Analyses on the effects of implementing implicit grid losses in the Nordic CCR, April 2018
22
EL
• 2018: price difference Sweden-Denmark has been zero for >2400 hours (54%)
• 2017: more than 5300 hours (61%)
• Total cost of losses during those5300 hours was approx. 0.9 M€ for a single HVDC line
An example: Kontiskan
23
Area 1
Area 2
Area 3
=~
=~
Flow-based market coupling
3-area IEEE RTS 72-bus SYSTEM
M ARKET CLEARING
Generator 1 max profit s.t. product ion const raints
TSO max profit s.t . transmission
constraints, HVDC loss factor
Consumer 2 max utility s.t. consumption const raints
Generator N max profit s.t. product ion const raints
Consumer M max utility s.t. consumption const raints
Generator 2 max profit s.t. product ion const raints
Consumer 1 max utility s.t. consumption const raints
Balancing constraints
• Mixed complementarity problem for the market clearing
24
LOOP FLOWS CONGESTIONNORMAL OPERATION
ECONOMIC LOSS 206.9 $/h
ECONOMIC BENEFIT 114.1 $/h
The penalization of the HVDC line results in an increase of losses in the AC system
With the introduction of the LF loop flows are avoided
Losses are covered by the price difference due to congestion
197.6 MW202.4 MW
271.6268.9
271.9271.9
ECONOMIC BENEFIT0 $/h
• Introducing an HVDC loss factor in the market clearing algorithm can have a positive or negative effect depending on the system under investigation
• Future work: Investigating different solutions to account for losses in non-radial HVDC systems
PowerlabDK at DTU• Power Hardware in the Loop Simulations:
– Robust control for varying inertia– Control of grid-forming HVDC in a zero-inertia AC grid– Coordinated control of HVDC for sharing reserves in
the Nordic region
• Implementation starting in September 2018
25
Development of a dynamic AC/HVDC Nordic model
26
+
• Danish system: already implemented in RTDS
• Nordic-44 system (Swedish equivalent, including Norway & Finland)– Adjusted system kinetic energy– Adjusted reactive power– Integration of wind power (20 GW of
wind by 2030)
• Working on an open-source VSC-HVDC converter model
27
Estimated kinetic energy in Nordic countries in 2020 and 2040
Development of a dynamic AC/HVDC Nordic model
Prior work on the Nordic Test System Development in
Thierry van Cutsem, Advancements in Power System Analysis Test Cases: Voltage Stability (18PESGM2383)Tue, Aug 7, 1:00pm-3:00pm, Room OC-D133+D134
Conclusions• multiDC: Holistic approach to the
emerging problems of multiple HVDC interconnections and grids– Robust control for varying inertia and
zero-inertia systems– Emergency power control coordination of
multi-area AC/DC grids– Market Integration of HVDC
• Developing solutions applicable to the North Sea Wind Power Hub
• Real implementation at PowerlabDK
28
Thank you!
www.multi-dc.eu
29
